Flight Safety Digest March 1992

Six Hours and 26 Minutes into a Four Hour, 40 Minute Flight

The stage is set for disaster when a weary international crew assumes that air traffic control is aware that a fuel emergency exists.

by Capt. Tom Duke Aviation Safety Writer

“On January 25, 1989, at approximately 2134 the flight crew executed a missed approach to eastern standard time, Avianca Airlines flight JFK. While returning to the airport, the air- 052 [AVA 052], a Boeing 707-321B with Colom- plane experienced a loss of power to all four bian registration HK 2016, crashed in a wooded engines and crashed approximately 16 miles residential area in Cove Neck, Long Island, from the airport. New York. [U.S.] AVA 052 was a scheduled international passenger flight from Bogotá, Colombia, to John F. Kennedy International Problems Accumulated Early Airport [JFK], New York, with an intermedi- ate stop at Jose Maria Cordova Airport, near The following is a brief reconstruction from Medellin, Colombia. Of the 158 persons aboard, the accident report of how the crew’s prob- 73 were fatally injured.” So began the accident lems accumulated. summary of the U.S. National Transportation Safety Board (NTSB) in its report of an aircraft • The captain ordered approximately 6,000 accident that would highlight numerous safety pounds of fuel, or a possible 30-45 min- issues. utes of extra flight, added prior to departure from Medellin. This raised the Because of poor weather conditions in the north- aircraft’s takeoff weight to the airfield eastern United States, AVA 052 was placed in performance limit. If the flight had ex- holding patterns three times by air traffic con- perienced no delays, it would have landed trol (ATC) for a total of approximately one at JFK with more than 20,000 pounds, hour and 17 minutes. During the third hold- or approximately two hours flight time ing pattern, the flight crew reported that the of fuel remaining. airplane could not hold longer than five minutes, that it was running out of fuel and that it • The weather data provided to the flight could not reach its alternate airport, Boston crew before departing Medellin was nine Logan International in Massachusetts. Subse- to 10 hours old. It indicated that JFK quently, after being cleared for an approach, had weather no worse than a 400-foot

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 1 ceiling, one mile visibility and winds • The crew of AVA 052 did not use avail- gusting to 25 knots with light rain. The able flight dispatch services to update alternate, Boston, however, was reported weather and alternate airport informa- as intermittently below legal minimums tion and might not have been aware of for filing as an alternate airport. the deteriorated weather conditions at JFK by the time they reached Norfolk. • Except for the short leg from Bogotá to Medellin, the captain had not flown pre- • The flight plan showed that AVA 052 viously with either the first officer or was early by approximately eight min- engineer. He chose to fly the leg to JFK. utes at Norfolk and had used up much According to the cockpit of the extra fuel loaded at Medellin. voice recorder (CVR), the This may have been the result of first officer made all the flying faster, heavier and at a lower radio calls in English and The crew … altitude than the original flight repeated air-ground com- might not have plan. After holding at Norfolk for munications to the captain 19 minutes before heading toward in Spanish. been aware of JFK, the aircraft departed 37,000 feet with 17,000 pounds of fuel at • The engines of the aircraft the deteriorated 1942 hours, approximately 20 min- had been modified with weather utes behind scheduled flight plan hush kits that, along with “top of descent” time. The required other factors such as age conditions … . 10 percent reserve fuel and extra and cumulative wear and 6,000 pounds was almost gone. tear on the aircraft’s aero- Without further delay, AVA 052 dynamics and powerplant could expect to land at JFK with efficiency, degraded fuel 14,500 pounds of fuel. This would consumption approximately 10 percent; be sufficient for about one hour and 30 this was considered in the fuel plan- minutes of flight at approach altitudes ning. Landing procedures for this air- and still leave slightly more than the craft call for the use of only 25 degrees 12,400 to 14,100 pounds required to go of flaps during landing for noise toler- to the Boston alternate and hold for 30 ance except when authorized for emer- minutes before running out of fuel. gency procedures. However, use of the autopilot during approaches with 25 • At 1943, near Atlantic City, N.J., at Boton degrees of flaps is not allowed below intersection, AVA 052 was held again, 500 above ground level (agl). Coupled for 29 more minutes. While holding at autopilot approaches with 40 to 50 de- Boton, the flight crew received several grees of flaps are certified to 62 feet agl altitude change clearances to 19,000 feet but because of the noise limitation, may and two expect further clearance (EFC) only be used in an emergency. changes. The crew requested information about Boston delays from Wash- • Aircraft records indicated recurring au- ington Center and were informed that topilot problems. A captain who had Boston was accepting traffic and that flown this aircraft on the flight just prior JFK would have to hold them “at most to the accident flight indicated there 30 more minutes.” Three minutes after was a problem with the flight director hearing that news at 2006, the AVA 052 in the approach mode. The NTSB re- crew informed ATC that they wanted port stated the belief that AVA 052 might to proceed to JFK, not Boston. At that have been flown manually from Colombia time, they might have had approximately and that the approach was flown with- 14,000 pounds of fuel remaining. out the aid of a flight director. By accepting the clearance to JFK with

2 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 the prospect of a 30-minute hold, the requested by ATC. Apparently assum- AVA 052 crew accepted the fact that ing they had asked for priority traffic they would not have enough fuel for handling, and therefore had declared any subsequent holding en route to Bos- an emergency, the Avianca flight crew ton. With no more than a 30-minute did not use the words “emergency” or hold, they could forecast about 4-5,000 “Mayday.” By now, they probably had pounds of fuel, less than 30 minutes of less than 10,000 pounds — less than an flight, on landing. At this point, the hour — of fuel remaining. decision to land at Kennedy was the least-worst choice The aircraft manufacturer suggests because, as the NTSB report landing the 707-321B with no less stated, the airplane did not “ … we think than 7,000 pounds, but some air- have sufficient fuel to fly to lines and the U.S. Federal Aviation its alternate airport. The flight we can’t do it Administration (FAA) treat that as crew never informed now we, we, informational and not a required Avianca’s contract dispatcher minimum amount. It is likely AVA at Kennedy of their fuel situ- don’t, we run 052 would now be landing with less ation or asked for assistance. out of fuel than 7,000 pounds of fuel remain- The weather at nearby Phila- ing even if there were no further delphia, Pa., closer than Bos- now.” delays. ton, was no lower than 700 foot ceilings and one half [Avianca’s route manual on the sub- mile visibility during this ject of flight crew and ATC proce- time period and for the re- dures for a minimum fuel situation mainder of the flight. requires pilots to: (1) Advise ATC of your minimum fuel status when your • AVA 052 then held at Camrn intersec- fuel supply has reached a state where, tion, 39 miles south of JFK, for 29 min- upon reaching destination, you cannot utes from 2018 to 2047. At 2031, it was accept any undue delay. (2) Be aware cleared to JFK and descended to 11,000 this is not an emergency situation but feet. However, because of missed ap- merely an advisory that indicates an proaches ahead of it, the aircraft was emergency situation is possible should sent back to Camrn at 2037. At 2044, any undue delay occur. (3) Be aware a without a new EFC, AVA 052 asked for minimum fuel advisory does not imply “estimates” and in the next minute ATC a need for traffic priority. (4) If the re- first cleared it to Kennedy, then gave it maining usable fuel supply suggests the an EFC of 2105. The AVA 052 crew then need for traffic priority to ensure a safer stated, “Well I think we need priority...” landing, you should declare an emer- and ATC asked them how long they gency, account low fuel, and report fuel could hold and what was their alter- remaining in minutes.] nate. At 2046, the crew told JFK they could hold five minutes that their al- • Approach control directed AVA 052 to ternate “was Boston but we think we slow to 180 knots and cleared it to Deer can’t do it now we, we, don’t, we run Park Intersection for a runway 22L ap- out of fuel now.” proach and to descend to 7,000 feet at 2048. The cockpit voice recorder (CVR) Twenty three seconds later, AVA 052 follows the scenario at 2053 for the next was cleared to Kennedy. It was told to 40 minutes to the crash. Because the switch frequency to Kennedy approach controller for Camrn did not hear part control and was in contact by 2047. No of the transmission about “we run out information about remaining fuel in min- of fuel now,” he did not pass that criti- utes was offered by the flight crew or cal information to approach control. The

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 3 use of the term “priority” by the Avi- Manual elevator trim inputs were heard anca first officer did not key a thought on the CVR transcript. On final approach, process connoting “critical situation” the crew had problems maintaining air- or “emergency” to ATC. Approach control speeds requested by the tower control- gave AVA 052 routine handling while ler. The final approach was very un- the flight crew members apparently be- stable. The captain used 50 degrees of lieved they were known to be needing flaps. That setting causes higher fuel priority treatment, or no further delays. burn but the aircraft is easier to handle They were given another six- to eight- than with the normal 25-degree flap set- minute delaying vectored circle and a ting. At three miles from the runway, long 15-mile final approach leg, but did the aircraft rapidly went above glide- not complain. The flight engineer briefed path and then well below, possibly from for an emergency low-fuel missed ap- a windshear. At 2.3 miles from the run- proach procedure as required when there way, with the ground proximity warn- is less than 1,000 pounds ing system (GPWS) repeatedly of fuel per main tank (4,000 sounding “whoop whoop pull up,” pounds total) remaining. the crew leveled the aircraft at 200 feet and began looking for the The first officer repeated all The CVR runway lights. ATC instructions to the cap- conversation tain in Spanish; the captain At 2123:23 the captain asked, “The apparently had difficulty contained runway, where is it?” hearing him because he re- ample evidence quested several repeats. The At 2123:27 the first officer re- CVR conversation contained of fatigue and sponded, “I don’t see it, I don’t ample evidence of fatigue see it.” and stress, and indications stress … . of flight director and ILS At 2123:28, after eight-seconds of radio tuning and setting flying 200 feet above the popu- problems. The engineer was lated area below, they began a asked to set power using rpm setting, missed approach .8 mile short of the an indication that the captain might have missed approach point. The urgency of had a problem with throttle alignment the GPWS alarms added to the crew’s and insufficient time to cross check power other stresses and could have made them settings. anxious to expedite a missed approach, which decreased their opportunity to The Automatic Terminal Information Ser- see the runway environment. There was vice (ATIS) recording indicated “200 sky no communication from the control tower obscured visibility 1/4 mile” and a 17- concerning their dangerously low alti- knot, slightly left crosswind. [Parallel tude and below-glideslope condition. runway 22R was below its 250-3/4 minimums.] The runway visual range (RVR) • From missed approach to fuel exhaus- was 6,000 feet (more than the 2,000 feet tion was approximately 10 minutes. required), and a moderate 10-knot wind- During the initial stage of the missed shear was reported to an aircraft closely approach, the captain told the first offi- following AVA 052. However, the lat- cer, “Tell them we are in emergency.” est RVR and the windshear report were not directed to AVA 052 by the local The first officer then told ATC, “ … control tower operator. we’ll try once again we’re running out of fuel.” • There are no conversations on the CVR indicating that the autopilot was in use. The tower responded, “Okay.”

4 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 The captain then asked the first officer, He incorrectly responded “one hundred “What did he say?” eighty” upon inquiry of the captain about the new heading. The first officer repeated his own transmission to the tower in Spanish. The captain again challenged the first officer, “Did you already advise that The captain again said, “Advise him we don’t have fuel?” we are emergency.” Then he asked, “Did you tell him?” The first officer responded in Spanish, “Yes sir, I already advise[d] him hun- The first officer, possibly preoccupied dred and eighty on the heading we are with a windshear report being given to going to maintain three thousand and another aircraft on the ra- he’s going to get us back.” dio, responded, “Yes, sir. I The captain said, “Okay.” already advised him.” The new The word “emergency” or controller gave • The flight had proceeded south, “Mayday” was still not stated away from the airport, for more or transmitted by the first no indication than a minute before approach con- officer. The crew was told that he was trol gave AVA 052 a new heading to continue the left turn to to 070 and transmitted “I’m gunna 150 degrees. aware of the bring you about fifteen miles north urgency of the east and then turn you back onto • After vectoring the aircraft the approach is that fine with you away from the field, at situation … . and your fuel?” 2124:39 the tower switched AVA 052 to approach con- The first officer responded, “I guess trol. When changing frequency to ap- so thank you very much.” proach control, the first officer reported “… we’ve missed … and we’re main- The Captain asked, “What did he say?” taining 2,000 ...” The flight engineer said, “The guy is The new controller gave no indication angry.” that he was aware of the urgency of the situation and did not inquire if they The first officer said, “Fifteen miles in were in an emergency status as he cleared order to get back to the localizer.” The them to climb to 3,000 feet. time was 2126:47, less than six minutes to fuel exhaustion. The captain again ordered the first officer, “Advise him we don’t have fuel.” • At 2129:11, when they were approximately abeam the outer marker, AVA The first officer radioed, “Climb and 052 asked approach control, “Can you maintain 3,000 and ah we’re running give us a final now?” out of fuel.” Approach responded, “Affirmative sir Approach responded, “Okay, fly head- turn left heading 040.” This new heading ing zero eight zero.” was the reciprocal of the landing runway and would take the aircraft for a long The first officer read back the clearance 15-mile final, not a shortened approach. and told the captain, “Three thousand A minute later at 2130:14, approach cleared feet, please” without advising the cap- another aircraft for a “left turn two five tain of the heading change. zero and … cleared for ILS.”

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 5 Avianca’s first officer read back the clear- • At 0234:00 the final controller called, ance as if it had been meant for AVA “Avianca zero five two you have uh 052. The captain made the turn, left al- you have enough fuel to make it to the titude and then AVA 052 was told to airport?” It was too late. “… climb and maintain 3,000 feet” by approach control. • At 2135, just after AVA 052 had disappeared from radar, a special weather At 2130:36, the first officer told approach, observation indicated improvement to “Negative sir we just running out of 300 foot ceiling and 3/4 mile visibility, fuel...” RVR (runway visual range) 5,500-6,000 feet. Approach then turned AVA 052 to a heading of 310 degrees, perpendicular to fi- • In retrospect, the flight might have been nal course, then to a heading of 360 de- able to receive a shortened approach grees with the explanation that “you’re and perhaps have returned after the number two for the approach missed approach had the flight crew I just have to give you enough openly stated they had a fuel emer- room so you make it without gency, used the word “Mayday” or … having to come out again.” “We just … refused the climbs and vectors. New lost two York’s controllers never became At 2132:11, approach turned aware of the extreme urgency, did AVA 052 to 330 degrees and engines and … not volunteer to ask the flight crew seconds later the engine we need if they had an emergency fuel situ- flameouts began. At 2132:49, ation or ask if they had enough fuel the first officer reported, “We priority until they heard the word “flame- just … lost two engines and please.” outs.” “We’re running out of fuel … we need priority please.” sir,” did not elicit a sense of extreme urgency on the part of the Approach turned them left to 250 de- controller, according to the taped con- grees to intercept the localizer and at versation. 2133.04 stated, “Avianca zero five two heavy you’re one five miles from the • This accident reflects the combination outer marker maintain two thousand of multiple events that eventually be- until established on the localizer cleared come overwhelming and result in a di- for ILS two two left.” saster. There was a long string of air- ground miscommunications and unveri- The first officer responded, “Roger, fied misunderstandings, poor flight plan- Avianca.” It was their last transmis- ning decisions, poor crew coordination, sion. The CVR tape’s power was lost 12 poor company dispatch and avionics seconds later. support procedures and rules, worse than forecast weather and an unexpected In very low ceilings, driving rain and number of delays. The result was ex- gusty winds during darkness, AVA 052 treme crew stress and fatigue that led glided to a crash landing in a wooded to inadequate communications and a neighborhood. The upslope impact in poorly executed approach and missed trees was gentle enough for 85 of the approach circuit … and ultimately fuel 158 persons on board to survive the exhaustion. impact forces. From the appearance of the cabin, none of the passengers was AVA 052 had been a challenge from the begin- warned of the impending crash. There ning; a four hour 40 minute flight ended six was no last minute electrical power avail- hours and 26 minutes later on a wooded hill- able to the cabin speaker system. side. As the report succinctly states, “If the flight

6 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 crew of AVA 052 had been able to complete the “breakouts,” or ATC-directed missed approaches. first ILS approach and land successfully, the The problem was not only one of maintaining accident would not have occurred … .” aircraft spacing in sequence for aircraft arriving from many directions but one of fitting missed approach aircraft back into the traffic Background Details Set Scene pattern. This caused further delays that resulted in extra holding times for others fol- The goal of the AVA 052 flight crew was to lowing in sequence. For example, the aircraft complete a safe, on-time flight to JFK. Their landing just ahead of AVA 052 had been worked scheduled arrival time was during busy inter- back into the pattern after an ATC-directed national arrival hours in the early evening when missed approach, when the captain expressed many domestic feeder airlines are also mak- concern about low fuel; he successfully made ing connections. As part of the U.S. National his second approach using the autopilot. Airspace System (NAS), a Central Flow Con- trol Facility (CFCF) located at FAA Headquar- The Avianca flight was to become one of 22 ters in Washington, D.C., attempts to manage missed approaches that afternoon and evening the flow of traffic into especially busy airports at JFK. At least 23 other flights were diverted throughout the country such as JFK. Domestic to other airports. Each missed approach ac- flights headed for airports experiencing land- counted for a three- to an eight-minute extra ing delays are held by CFCF on the ground at hold for other aircraft heading to JFK. The the U.S. departure airports until the threats of NTSB report did not mention how long other inflight holding dissipate. This is a fuel-sav- arriving flights were held or delayed. ing procedure and an excellent safety practice. Foreign arrivals, however, are not subject This unanticipated situation caused AVA 052 to CFCF ground delays and are fed to hold three times beginning at into the system while airborne. Norfolk for a total of one hour, 17 minutes plus another eight-minute Because the weather situation at JFK The Avianca vectored circle after the crew told on that January day became worse controllers they were then so low than predicted and the wind con- flight was to on fuel that they could not go to ditions and low ceilings did not become one of their alternate. The ATC controllers allow the use of two runways for were busy with rapidly changing situ- landings, the expected airport ac- 22 missed ations and they did not hear every- ceptance rate of 33 aircraft arrivals approaches … . thing that flight crews were trans- per hour became drastically reduced mitting, or pass all information on and arrivals began backing up. Rout- to the next controller in the sequence. ing for AVA 052 placed the aircraft Flight separation requirements var- on an arrival route from the south where it ied from 20 miles en route to five-miles on was placed in a trail of other traffic and held final. There was little time for interruptions or with many domestic carriers originating from extra clarification by air traffic controllers. points south of Norfolk, Virginia. As a result of this hectic situation, by the time The main impediment to AVA 052 arriving on of their first approach to Kennedy, the flight time was excessive holding caused by earlier crew of AVA 052 had logged six hours, 17 min- missed approaches by other aircraft at JFK due utes on a planned four hour, 40-minute flight. mainly to difficulties pilots were having with The NTSB stated that it believes that AVA 052 windshear, low ceilings and visibility. Air traffic and its dispatcher should have kept each other controllers were also having trouble keeping more informed before the emergency situa- proper aircraft separation because there were tion developed. strong and varying tailwinds of approximately 70 knots on the downwind leg and headwinds In order to make a decision to go to an alter- of 30 knots or more on final that were forcing nate when the destination is at or near mini-

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 7 mums, one of the facts a captain wants to know convey fuel status and the need for special is when to expect to land. The only hard infor- handling; mation the AVA 052 flight crew was able to obtain was EFC times which ATC must give • ATC flow control procedures and re- crews in the event of lost communications. sponsibilities to accommodate aircraft Because of the dynamics of the situation that with low fuel states; and, evening, an expected landing time could not be predicted by either the crew or ATC. One of • Flight crew coordination and English the variables was the number of aircraft ahead language proficiency of foreign crews. of AVA 052 that would miss their approaches and how many of them that would proceed to Recommendations concerning these issues were their alternate airports. The flow plan did not addressed to the FAA and the Director, adequately consider extra approaches. Departmento Administrativo de Aeronautico Civil (DAAC), Colombia. Probable Cause Determined Conclusions Reviewed The NTSB determined that the probable cause of this accident was the failure of “the flight It is appropriate for this discussion to review crew to adequately manage the airplane’s fuel all of the 24 findings listed in the NTSB re- load, and their failure to communicate an emer- port. In parentheses following some findings gency fuel situation to air traffic control be- are the person(s) and main category of hu- fore fuel exhaustion occurred.” Contributing man performance factor or error. The find- to the accident according to the board, was ings are below: “the flight crew’s failure to use an airline operational control dispatch system to assist them 1. The accident occurred when the airplane’s during the international flight into a high-density engines lost power as a result of fuel exhaus- airport in poor weather. Also contributing to tion while the flight was maneuvering for a the accident was inadequate traffic flow man- second instrument approach to JFK airport. agement by the [U.S.] Federal Avia- tion Administration [FAA] and the 2. Examination of the airplane revealed lack of standardized understandable … failure … no malfunction of the engines or fuel terminology for pilots and control- to adequately system components that could have lers for minimum and emergency caused a premature fuel exhaustion. fuel states.” manage the airplane’s fuel 3. The flight crew was not provided The board also determined that wind- with, and they did not request be- shear, crew fatigue and stress were load, and … to fore departure, the most current factors that led to the unsuccessful communicate weather forecast available for the completion of the first approach, and destination or selected alternate air- thus contributed to the accident. an emergency port. (captain, flight planning/dis- fuel patcher, flight planning) Numerous safety issues were raised in the NTSB report. They include: situation … . 4. The alternate airport selected for the flight at the time of departure • Pilot responsibilities and dispatch re- did not meet the prescribed weather criteria sponsibilities regarding planning, fuel for an alternate based on weather information requirements and flight following dur- provided to the crew at the time of departure. ing international flights; The weather conditions worsened at both the destination and alternate while the flight was • Pilot to controller communications re- en route. (captain, flight planning/airline dis- garding the terminology to be used to patcher, flight planning)

8 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 5. The flight plan for AVA 052 did not reflect (captain, communications/first officer, com- the most current upper air data or the actual munications) gross weight of the aircraft upon departure from Medellin. (captain, flight planning/dis- 13. The controllers’ actions in response to AVA patcher, flight planning) 052’s requests were proper and responsive to a request for priority handling. They did not 6. The flight crew had received appropriate understand that an emergency situation ex- flight and ground training for the isted. (ATC, communications) flight, and they possessed appropriate flight and medical certifica- 14. The first officer, who made all tion required by the government recorded radio transmissions in En- of Colombia. The first glish, never used the word “emer- officer … never gency,” even when he radioed that 7. The flight crew was experienced two engines had flamed out, and in conducting Boeing 707 flights used the word he did not use the appropriate from Colombia to the United States. phraseology published in U.S. aero- “emergency” … . nautical publications to communi- 8. There was no flight following or cate to air traffic control the flight’s interaction with the Avianca Air- minimum fuel status. (first officer, lines dispatcher for AVA 052 fol- communications) lowing takeoff from Medellin. None was required by the airline’s operations specifications 15. The weather conditions at JFK were worse issued by the FAA under 14 Code of Federal than forecast. (weather/ATC, communications) Regulations (CFR) 129, the U.S. regulation that covers non-U.S. air carriers operating into the 16. The captain did not fly the ILS approach in United States. (dispatcher, flight planning/DAAC a stabilized manner, which led to a serious rulemakers, organizational oversight/airline deviation below the glideslope and to his ini- operations, organizational oversight) tiation of a go-around. (captain, procedural behavior) 9. There is no record that while en route the flight crew requested updated weather infor- 17. A windshear on the approach path may have mation from any source regarding the desti- contributed to the captain’s poor performance nation or alternate airport. (captain, flight plan- on the ILS approach. Although other flights suc- ning) cessfully completed the approach through the same wind conditions, the captain’s performance 10. The flight crew did not adequately com- on the approach was probably degraded by fa- municate the increasingly critical fuel situa- tigue after the long flight and by his reliance on tion to the controllers who handled the flight. raw glideslope position data rather than on au- (captain, communications/first officer, com- topilot or flight director guidance. (captain, physi- munications) ological/airline maintenance, organizational oversight/captain, decision making) 11. The first officer, who made all recorded transmissions to U.S. controllers, was suffi- 18. The FAA traffic management programs failed ciently proficient in English to be understood to effectively manage the traffic volume at JFK, by air traffic control personnel. leading to excessive delays and airborne hold- ings, including more than one hour for AVA 12. The first officer incorrectly assumed that 052. ( FAA, organizational oversight) his request for priority handling by air traffic control had been understood as a request for 19. The FAA’s traffic management programs emergency handling. The captain experienced for JFK did not adequately account for over- difficulties in monitoring communications seas arrivals and missed approaches at JFK. between the flight and air traffic control. (FAA, organizational oversight)

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 9 20. Cabin crew members and passengers were NTSB’s 24 findings uncovers 17 findings that not warned of the impending crash landing, directly or indirectly identify 31 individual which may have contributed to the severity of human performance errors by the flight crew, the injuries sustained. (captain, procedural air traffic controllers, weather services, gov- behavior/second officer, procedural behavior/ ernment oversight agencies, company dispatch, senior flight attendant, procedural behavior) maintenance and management (Table 1).

21. The serious and fatal injuries were the re- The parties to the investigation, especially Avi- sult of blunt force trauma because of high ver- anca Airlines, were all very cooperative dur- tical and longitudinal deceleration forces during the year-long investigation. A public hearing the impact sequence. ing held in Long Island, N.Y., in June 1990, brought out many of the international airline 22. The emergency evacuation slides were inop- pilots’ and human factors research community’s erative because of the lack of appropriate floor assessments of the problems highlighted in attachment hardware. (DAAC, organizational this accident. However, because the flight crew oversight/maintenance, organizational oversight/ did not survive the accident, an in-depth hu- senior flight attendant, procedural behavior) man factors investigation was not possible to help explain why the flight crew’s decision- 23. There were no shoulder harnesses or iner- making process broke down. Stress and fa- tia reels installed on captain’s and first officer’s tigue are illusive causal factors based on in- seats. (company, organizational oversight) tuitive information and the experiences of the past. How the flight crew got into this tragic 24. The response of fire and rescue personnel situation and what was happening on the ground was timely and effective, and the use of heli- are not likely to be fully understood. copters by the Nassau County Police Depart- ment probably saved lives. Recommendations Reveal Concerns, Offer Solutions Human Factors Were Significant The NTSB made several recommendations as This accident investigation is a human perfor- a result of this accident investigation, both to mance investigator’s cornucopia. A scan of the the FAA and to the DAAC of Colombia.

Table 1 Summary of Human Performance Errors

Flight Communi- Procedural Physio- Decision Organization Person (Total) Planning cations Behavior logical Making Oversight

Capt (10) 4 2 2 1 1 F/O (4) 4 F/E (1) 1 Sen F/A (2) 2 Dir Maint(2) 2 Dir Ops (2) 2 Dispatch (4) 4 DAAC (2) 2 ATC Cont (1) 1 Weather (1) 1 FAA Mgt (2) 2

TOTALS (31) 8 8 5 1 1 8

10 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 Recommendations to the FAA situation is occurring and offer assistance. [These recommendations were • Develop in cooperation with the Inter- published by Flight Safety Foundation, national Civil Aviation Organization Accident Prevention Bulletin, April 1990.] (ICAO) a standardized glossary of defi- nitions, terms, words, and phrases to Recommendations to the DAAC, Colombia be used that are clearly understandable to both pilots and air traffic controllers • Review policies, procedures, training regarding minimum and emergency fuel and oversight activity to ensure that communications. adequate emphasis is being placed on the dual responsibility that flight dis- • Conduct a comprehensive study of the patchers and flight crews have in keeping central flow control facility and the traffic each other informed of events and situ- management system, by the agency’s ations that differ from those mutually Office of Safety/Quality Assurance, to agreed upon in the dispatch release. determine the effectiveness and appro- priateness of training, responsibilities, • Require that Avianca Airlines incorpo- procedures and methods of application rate cockpit resource management (CRM) for the traffic management system. and line oriented flight training (LOFT) concepts into its flight crew training • Require that transport category airplane program. flight manuals include procedures speci- fying minimum fuel values for various phases of airline flights at which a landing Report Included Dissents should not be delayed and when emergency handling by ATC should be re- In a dissenting statement, NTSB Member Jim quested. The manual requirement and Burnett supported the causes and recommen- associated amendments to regulations dations, but listed four specific unsatisfactory and procedures should include criteria services provided to AVA 052 in holding, weather for when ATC must be notified that the information, passing on the fuel situation to airplane must be en route to its desti- other controllers and updating the JFK ATIS. nation or alternate airport via routine He further stated that the FAA allowed more handling, and when emergency handling aircraft into the system than it could safely is required. handle.

• Incorporate into air route traffic con- Member Christopher Hart stated that the Avi- trol centers equipment to provide a re- anca crew should have known to use the words corded broadcast of traffic management “Mayday” or “emergency” to communicate their information that can be monitored by dangerous situation to ATC and that a lack of all aircraft within each center’s bound- standardized terminology is not a contribut- aries to provide pilots with early indi- ing factor. cations of potential delays en route. Also, the DAAC stated that there were many • Prior to the release of the final accident inadequacies and ambiguities on the part of report, the NTSB had made three ear- U.S. ATC procedures that misled the Avianca lier safety recommendations to the FAA. flight crew into thinking they were receiving These concerned notifying all domestic special handling when they were not. It stated and foreign air carriers to be familiar that they were given holding after telling ATC with the U.S. National Airspace Sys- they could not make their alternate and were tem and its rules and procedures, and given a normal long landing pattern after tell- for air traffic controllers to request clari- ing ATC they were running out of fuel. DAAC fication whenever a possible emergency also stated that EFC times should include in-

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 11 formation on when a pilot can expect to land. Studies Division of the U.S. National Transportation The agency said ATC should require an active Safety Board (NTSB). He is a former captain with flight following system with landing delay in- Independent Air and flew international charters formation for international flights not subject in the Boeing 707. to ground holds. o Duke spent 30 years in the U.S. Air Force. He was director of safety of the Air Force Reserve, and managed About the Author a worldwide investigative and accident prevention program that won several awards. He has more than Capt. Thomas A. Duke is a free-lance aviation 11,400 hours of military and airline flying and more safety writer. He previously was with the Safety than 16 years of active safety management.

Aviation Statistics

Worldwide Passenger Traffic and Air Carrier Safety Calendar Year 1991 by Shung C. Huang Statistical Consultant

Worldwide Passenger Traffic the hostilities ended in the spring of 1991, it Declines was expected by most industry observers that there would be a resurgence in worldwide air travel. However, even the quick end of the The Persian Gulf War and the threat of greatly war, followed by peace negotiations between increased terrorism influenced a slowdown in the Arab States and Israel, failed to stimulate worldwide passenger air traffic in 1991. When a favorable environment for worldwide pas-

12 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 senger traffic. On the contrary, the aftereffects Passenger-kilometers Performed by of the war, including the political stalemate in Worldwide Airlines the Gulf area, the standstill in the world eco- Calendar Year 1881-1991 nomic recovery and continuing concerns about the potential effects of terrorist attacks upon 2000000 the safety of flight continues a severe decline 1800000 in both international and U.S. domestic air 1600000 travel. 1400000 1200000 As a result of the negative forces, the world- 1000000 wide air carrier industry experienced its first 800000 decline in passenger traffic in two decades dur- 600000 ing 1991. Preliminary estimates from the Inter- 400000 national Civil Aviation Organization (ICAO) 200000 indicate that worldwide total passenger and 0

Worldwide Passengers-kilometers (Million) cargo ton-kilometer figures will show a decline Worldwide Passengers-Kiilometers (Million) 1981 1988 1989 1990 1991 of four percentage points in 1991 compared to 1982 1983 1984 1985 1986 1987 Source: ICAO Year the previous year. The annual distribution of worldwide airline passenger-kilometers for the Figure 1 past 10 years is shown in Figure 1. airlines for the first 11 months dropped to 418 million from 436 million in 1990, a decline of Decline in U.S. Passenger Traffic 4.5 percent; the passenger-miles flown dropped from 433 billion to 422 billion, a reduction of The decline of passenger traffic in the United 2.7 percent. The monthly comparison of pas- States, which accounts for more than 40 per- sengers carried and passenger-miles flown by cent of worldwide passenger traffic, was even U.S. airlines for 1990 and 1991 is shown in more severe than the international decline. In Table 1. Note that during the 11-month period 1991, the number of passengers carried by U.S. for which information is available, in not a

Table 1 Passengers Carried and Passenger-miles Performed U.S. Airline all Operations Scheduled and Non-scheduled Services 1990 vs. 1991 Passengers Carried (000) Passenger-miles (Million) Month 1990 1991 Change 1990 1991 Change January 35,548 34,897 -1.8 35,153 34,828 -.09 February 34,790 32,063 -7.8 32,965 29,610 -10.2 March 41,996 37,637 -10.4 39,993 36,142 -9.6 April 39,560 37,803 -4.4 37,987 36,951 -2.7 May 39,492 38,623 -2.2 38,419 38,789 +1.0 June 42,419 40,334 -4.9 42,819 41,675 -2.7 July 43,618 42,885 -1.7 45,770 45,269 -1.1 August 46,316 44,677 -3.5 48,763 48,104 -1.4 September 37,073 36,010 -2.8 38,173 37,578 -1.6 October 39,599 38,426 -3.2 39,051 38,925 -0.3 November 37,525 35,614 -5.0 35,699 34,367 -2.7 December 37,330 not available 37,331 not available January-November Total 436,525 418,971 4.5 433,792 422,238 -2.7

Source: Air Carrier Traffic Statistics Monthly, Calendar Year 1990 and 1991 Research and Special Programs Administration, U.S. Department of Transportation.

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 13 Table 2 U.S. Accidents, Fatalities and Rates Air Carriers and General Aviation 1991 (Preliminary Data) Accident Rates Per 100,000 Per 100,000 Accidents Total Aircraft Aircraft Hours Departures Total Fatal Fatalities Hours Flown Departures Total Fatal Total Fatal

Air Carriers Operating Under 14 CFR 121

Scheduled 26 4 62# 11,250,000 7,500,000 0.231 0.036 0.347 0.053 Nonscheduled 1 0 0 580,000 270,000 0.172 0 0.370 0

Air Carriers Operating Under 14 CFR 135

Scheduled 22 8 99# 2,100,000 2,700,000 1.048 0.381 0.815 0.296 Nonscheduled* 84 26 69 3,270,000 n/a 2.57 0.80 n/a n/a

General Aviation+ 2,143 414 746 30,760,000 n/a 6.90 1.35 n/a n/a

Exposure data estimate source: U.S. Federal Aviation Administration (FAA).

# Both of these fatality totals include the 12 persons killed aboard a Skywest commuter aircraft (Scheduled 14 CFR 135) and the 22 persons killed aboard the USAir airliner (Scheduled 14 CFR 121) after the two aircraft collided.

* Accidents on non-U.S. soil and in non-U.S. waters are excluded.

+ Includes accidents involving U.S. registered civil aircraft flown under rules other than 14 CFR 121 and 14 CFR 135. Accidents on non-U.S. and in non-U.S. waters are excluded.

n/a Data not available.

Worldwide Airline Fatal Accidents and Fatalities

30 2000 1800 25 Ten-year average 1600 20 1400 1200 Ten-year 15 average 1000

Fatalities 800 10 600 Fatal Accidents 5 400 200 0 0 1988 1989 1990 1991 1981 1982 1983 1984 1985 1986 1987 1981 1988 1989 1990 1991 Year 1982 1983 1984 1985 1986 1987 Year Source: FSF Flight Safety Digest 1981-1990

Figure 2

14 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 single month in 1991 was the passenger traffic operations could be off by as much as 50 per- higher than that in the corresponding month cent for that month. of 1990. Airline Safety Figures Air Traffic Declines in Reflect A Good Year The Former Soviet Union During 1991, worldwide airlines operating large The Mideast crisis might not have caused sig- aircraft were involved in 13 fatal accidents nificant negative effects on former-Soviet air that accounted for 760 fatalities. Fourteen jet passenger traffic, but news reports have noted transport aircraft used by worldwide airlines that air passenger traffic in 1991 worsened in were totally destroyed. Figure 2 shows the an- grim parallel with the deterioration of U.S.S.R. nual numbers of fatal accidents and fatalities political and economic systems during the tran- involving worldwide airlines for the period sition to the Commonwealth of Independent 1981-1991. It shows that in 1991 worldwide States (CIS). As a result of a decreasing supply airlines recorded fewer fatal accidents and fa- of fuel from the state energy monopoly, an talities than those in most prior years, and estimated 40 percent of Aeroflot planes were fewer than the average of the past 10 years. To grounded at one point. In late December 1990, be exact, the average for fatal accidents is 20 the transport minister reported that Aeroflot and that for fatalities is 989 as indicated on

Table 3 Worldwide Airline Fatal Accidents and Jet Transport Hull-losses Calendar Year 1991

Date Location Aircraft Damage Fatalities Phase Remark

1/2 Los Angeles, Calif., U.S. B-737 Destroyed 34 Takeoff Collided with other aircraft on runway after landing 2/17 Cleveland, Ohio, U.S. DC-9 Destroyed 2 Takeoff Crashed out of control 2/20 Puerto Williams, Chile BAe 146 Destroyed 20 Landing Overran the runway and crashed into sea 3/3 Colorado Springs, Colo., U.S. B-737 Destroyed 25 Approach Crashed on approach loss of control 3/5 Trujillo, Venezuela DC-9 Destroyed 43 Approach Crashed into high ground in low visibility 3/12 New York, N.Y., U.S. DC-8 Destroyed none Takeoff Aborted takeoff, rupturing fuel line. Fire after impact. 3/23 Tashkent, Russia An-24 Destroyed 31 Landing Veered off runway and crashed into concrete blocks 5/23 St. Petersburg, Russia TU-154 Destroyed 10 Landing Crashed on Landing 5/26 Suphan Buri, Thailand B-767 Destroyed 223 Cruise Loss of control; malfunction of thrust-reversers 6/26 Sohoto, Nigeria BAC 1-11 Destroyed 3 Landing Crashed on emergency landing in adverse weather 7/11 Makhackala Yak-40 Destroyed 34 Approach Crashed into high ground on approach 8/16 Imphal, India B-737 Destroyed 69 Approach Crashed on approach 11/7 Jeddah, Saudi Arabia DC-8 Destroyed 261 Takeoff Crashed on emergency landing shortly after takeoff 12/7 Tripoli, Libya B-707 Destroyed none Takeoff Crashed on takeoff run 12/27 Stockholm, Sweden MD-81 Destroyed none Takeoff Engine failure/crashed on emergency landing 12/29 Shienchu, Taiwan B-747 Destroyed 5 Approach Crashed on emergency landing due to engine failure

Source: News reports compiled by author.

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 15 Figure 2 by horizontal lines. Note that the an- Former-Soviet Airline Safety nual distribution of fatal accident and fatali- Statistics Show Short-term Decline ties appears to be random. Because not all aviation accident information involving former-Soviet airlines is reported, U.S. Airline Safety Figures complete statistics are difficult to obtain. Ac- Are Mixed cording to news reports, aviation safety in the Soviet Union eroded badly during the past In the United States, the National Transpor- three years. The news media has reported that tation Safety Board (NTSB) reported that U.S. Aeroflot, the ex-U.S.S.R. state airline, was in- airlines operating large aircraft in 1991 were volved in 36 accidents in 1991 that accounted involved in 27 accidents, four of them fatal for 252 fatalities compared to 27 fatal acci- that resulted in 62 fatalities. This represents dents and 194 fatalities in 1990. two fatal accidents fewer than, but 23 fatalities more than, those recorded in 1990. The NTSB also reported that commuter air carri- Worldwide Airline Fatal ers were involved in 22 accidents, eight of Accidents Listed for 1991 which were fatal, that accounted for 99 fatalities. In terms of fatalities, 1991 was the Table 3 is a listing of fatal accidents and jet worst safety year ever for commuter air car- transport hull-losses involving worldwide air- riers. The NTSB accident statistics for U.S. lines that operated large aircraft during 1991. air carriers, commuter air carriers, air taxis It includes three fatal Russian accidents of the and general aviation for 1991 are shown in 36 accidents reported by Aeroflot. Informa- Table 2. tion in table 3 is preliminary. o

Reports Received at FSF Jerry Lederer Aviation Safety Library

Reference

Updated Reference Materials (Advisory Circulars, U.S. FAA):

Numbers(s) Mo/Yr Subject 43-4A July 1991 Corrosion Control for Aircraft (Cancels AC 43-4 dated May 1973) 150/5200-30A Oct 1991 Airport Winter Safety and Operations (Cancels AC 150/5200- 30 dated April 1988) 150/5200-30A Nov 1991 Change 1 to Airport Winter Safety and Operations 150/5345-28D Nov 1991 Change 1 to Precision Approach Path Indicator (PAPI) Systems 150/5345-42C Oct 1991 Change 1 to Specifications for Airport Light Bases, Transformer Housings, Junction Boxes, and Accessories

16 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 National Transportation Safety Board (U.S.) Safety Recommendations:

Number(s) Mo/Da/Yr Subject A-91-104/121 12/03/91 USAir Boeing 737-300 and Skywest Fairchild Metro-liner collision, Los Angeles, California, U.S., February 1, 1991

New Reference Materials: Reports Advisory Circular 25-17, 11/25/91, Announce- ment of Availability of Advisory Circular 25- FAA Registry Systems: Key Steps Need to be Per- 17, Transport Airplane Cabin Interiors Crash- formed Before Modernization Proceeds. Report to worthiness Handbook. — Washington, DC : the Chairman, Permanent Subcommittee on U.S. Federal Aviation Administration, 1991. Investigations, Committee on Governmental 1 p. Affairs, U.S. Senate/ United States General Accounting Office. — Washington, D.C. : Gen- Summary: Advisory Circular 25-17 provides eral Accounting Office**, [1991]. Report GAO/ acceptable certification methods, but not nec- IMTEC-91-29. 13 p. ; 28 cm. essarily the only acceptable methods, of dem- onstrating compliance with the crashworthi- Key Words ness requirements of Part 25 of the Federal 1. United States — Federal Aviation Admin- Aviation Regulations (FAR) for transport cat- istration. egory airplanes. This announcement pro- 2. Airplanes — Inspection — United States. vides notice that Advisory Circular 25-17, date 3. Air Pilots — Drug Use — United States. July 15, 1991, is available only as a sale docu- 4. Air Pilots — Registers — United States. ment from the Superintendent of Documents, and is not available from the U.S. Federal Summary: GAO reports on the Federal Avia- Aviation Administration. Copies may be or- tion Administration’s (FAA) modernization plans dered from: Superintendent of Documents, for its airmen and aircraft registry systems. The U.S. Government Printing Office, Washing- Anti-Drug Abuse Act of 1988 requires changes ton, D.C. 20402. Stock Number 050-007- to these systems to make them more effective 00915-1. The cost of AC 25-17 is $11.00 (U.S.) in serving the needs of law enforcement agen- for U.S. orders. Orders for mailing to for- cies involved in aviation drug interdictions and eign countries mustinclude an additional $2.75 investigations. GAO found that FAA has not (U.S.). adequately defined users’ needs for the system, leaving FAA with little assurance that it has selected the most appropriate technologi- cal solution, and increasing the risk of cost over- Advisory Circular 120-55, 10/23/91, Air Car- runs and schedule delays. At the conclusion of rier Operational Approval and Use of TCAS the GAO review, FAA officials stated that they II. — Washington, DC : U.S. Federal Aviation planned to correct the deficiencies GAO identi- Administration, 1991. 31 p. in various pagings. fied. [Modified Results]

Summary: This Advisory Circular (AC) provides an acceptable means, but not the only means, to address Traffic Alert and Collision The Prevalence of Aphakia in the Civil Airman Avoidance System (TCAS) issues related to Population. Final Report/Van B. Nakagawara, installation and use of TCAS II regarding com- Faredoon K. Loochan, Kathryn J. Wood (Civil pliance with Federal Aviation Regulations (FAR) Aeromedical Institute). — Washington, D.C.: Parts 121, 125, and 129 requirements for air Federal Aviation Administration, Office of Avia- carriers. [Purpose] tion Medicine; Springfield, Va., U.S.: Avail-

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 17 able through NTIS*, [1991]. Report No. DOT/ Key Words FAA/AM-91-14. v, 13 p.: charts; 28 cm. 1. Air Traffic Controllers — Selection and Appointment — United States. Key Words 1. Air Pilots — Certification — United States. Contents: Overview of National Airspace Hu- 2. Air Pilots — Medical Examinations — United man Resource Management Planning Process/ States. Earl Pence (Fu Associates) and Shelly A. Tho- 3. Aphakia. mas — Employing Air Traffic Controllers/Jay 4. Eye — Abnormalities. C. Aul (FAA) — Procedures for Selection of Air Traffic Control Specialists/Carol A. Man- Summary: Aphakia has become increasingly ning (Civil Aeromedical Institute) — Evalua- prevalent in the civil airman population. The tion Issues in the Selection of Air Traffic Con- FAA allows civilian airmen with aphakia to trollers/Hilda Wing and Darlene M. Olson (FAA) fly with waivered certificates. The increased — Discussion of Selection of Air Traffic Con- application and modification of surgical pro- trollers: Complexity, Requirements, and the cedures for cataract extraction, coupled with Public Interest/Phillip L. Ackerman (Univer- possible visual complications from these pro- sity of Minnesota). cedures in flight operations, strongly suggests continued specialized aeromedical certification Summary: The essays in this report represent and clinical research review. This study ana- presentations made at the 98th Annual Con- lyzes the distribution of aphakia in the civil vention of the American Psychological Asso- airman population by type (unilateral, bilat- ciation, August 10-14, 1990, in Boston, Mass., eral), class of airman medical certificate, and U.S. The presentations address the diverse gender for a four-year period (1982-1985). [Modi- process of valid selection for the highly de- fied author abstract] manding occupation of air traffic control specialists. Selection of Air Traffic Controllers: Complexity, Requirements, and Public Interest. Final Report/ *U.S. Department of Commerce edited by Hilda Wing (Federal Aviation National Technical Information Service (NTIS) Administration) and Carol A. Manning (Civil Springfield, VA 22161 U.S. Aeromedical Institute). — Washington, D.C.: Telephone: (703) 487-4780 United States. Federal Aviation Administra- tion, Office of Aviation Medicine; **U.S. General Accounting Office (GAO) Springfield, Va., U.S.: Available through NTIS*, Post Office Box 6012 1991. Report DOT/FAA/AM-91/9. v, 38 p.: Gaithersburg, MD 20877 U.S. ill.; 28 cm. Telephone: (202) 275-6241

18 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 Accident/Incident Briefs

This information is intended to provide an aware- However, the close proximity of the cargo plat- ness of problem areas through which such occur- form on the aircraft’s left, the cargo building rences may be prevented in the future. Accident/ in front and a walkway support structure to incident briefs are based upon preliminary infor- its right resulted in tight quarters, with scrape mation from government agencies, aviation orga- marks from previous incidents evident on the nizations, press information and other sources. This cargo building. When the tug driver attempted information may not be accurate. to connect the towbar, he noticed that the aircraft’s nosewheels were slightly to his right AAirir C Carrierarrier of the centerline painted on the ramp, and at an angle to it, which combined with the close- ness of adjacent structures, would make any extra maneuvering of the aircraft very difficult. He left to telephone for help from the next ground crew shift because he was aware that there should have been three other ground crew members available to assist him with the pushback. Hurried Departure Goes Nowhere While the tug driver was away, the aircraft operator’s cargo manager arrived to check the Boeing 707: Moderate damage to fuselage. No aircraft’s documentation. He extended the cargo injuries. platform to within 18 inches of the fuselage to board it but, on leaving, did not retract or The cargo-configured jetliner was parked ad- lower the platform, because he saw a man jacent to a fixed cargo handling system de- connect the air start unit and then sit on the signed for use with that type aircraft. It had tug and assumed he would retract the loading been parked there at 1115 hours for unloading platform. and reloading prior to a scheduled 1330 departure. The captain was anxious to meet the revised departure time that was rapidly approaching, The unloading of the cargo was accomplished so he requested engine start and pushback clear- using a fixed, high-lift pallet loader on the left ance even though the next shift of ground support side of the aircraft that is able to be extended personnel had not arrived. One of the carrier’s or raised to be aligned with the cargo door of ground technicians assisted with engine start the aircraft. However, loading was delayed and removal of the ground power units and unexpectedly and the scheduled departure slot communicated with the pilot during start and of 1330 was extended two hours to 1530. Al- pushback. Neither the technician nor the tug though the aircraft was loaded in time for the driver accomplished a pre-pushback inspec- revised departure, the contracted loading crew tion of the ramp. members reached the end of their shift a short time before the new departure time and they The pushback was begun without the full ground departed, leaving the tug driver as the only crew and with no attempt to control vehicular ground crew member with the aircraft. The traffic on the service roads that were located flight crew was already aboard. to the rear and the side of the aircraft. After the aircraft had moved approximately five feet, The tug driver retracted and lowered the loading the captain sensed something was not right platform and positioned the tug for the pushback. and ordered the pushback stopped. A check

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 19 of the fuselage revealed that the aircraft had assumed, as had the ATC controller, that the struck the protruding cargo loading platform, Boeing would complete its parking procedure which had not been retracted after the company’s normally. cargo manager left the aircraft. During pushback of the Airbus, the pilot coor- The aircraft was out of service for 19 days dinated startup of the right engine with the while repairs were effected. ground engineer who then had to divide his attention between observing proper engine fan rotation on the right side of the airplane and Three Assumptions monitoring the left wingtip for clearance from Lead to Trouble the Boeing. (The Airbus was being pushed back in a turn to its right, so the left wingtip would Airbus A310: Substantial damage. No injuries. pass behind the tail of the Boeing.)

Boeing 757: Substantial damage. No injuries. As the Airbus exited its parking stand, the ground engineer observed that the left wing- The Boeing 757 had taxied for a scheduled tip was swinging around toward the tail of the departure, but a technical fault had been de- Boeing that had unexpectedly stopped part of tected and it had been cleared back to its park- the way out of the parking stand. He advised ing stand by air traffic control (ATC) to have the flight deck over the intercom and tried to maintenance attend to the problem prior to signal the tug driver, whose attention was fo- departure. cused on correctly positioning the aircraft’s nosewheel. The Airbus was being prepared for departure from a parking stand on the opposite side of a The left wingtip of the Airbus collided with taxiway from the parking stand to which the the underside of the Boeing’s tailplane and Boeing 757 was returning. Parking was nose- rear right fuselage, causing substantial dam- in for both aircraft, with pushback being pro- age to that aircraft and damage to the wingtip vided by tugs. and trailing edge of the Airbus. There was no fire and passengers of the aircraft were de- The ground controller observed that the Boe- planed without further incident by use of mo- ing 757 had passed behind the still parked bile stairways. Airbus, turned onto the centerline of its parking stand and was moving forward into its Investigation revealed that as the Boeing be- parking position. The parking stands were on gan to enter its parking stand and had been a cul-de-sac to which the controllers had lim- assumed by ATC and the Airbus ground crew ited vision and having seen the returning air- to be completing its parking normally, the flight craft moving into its parking stand, the con- crew realized that the entry guidance system troller assumed that the parking maneuver would was not illuminated and decided to stop the be completed normally. Therefore, he issued aircraft until the parking guidance system was a pushback clearance to the Airbus which was illuminated or a ground crew member arrived ready for departure. to provide guidance. The tail of the aircraft was protruding into the taxiway by approxi- The Airbus began its pushback, supervised by mately 69 feet. Because of heavy radio traffic, one ground engineer positioned on the left the flight crew was unable to inform ATC of side of the aircraft who was in interphone contact its predicament immediately. The aircraft was with the cockpit crew. The only other ground stopped only briefly until the call was made; crew member was the tug driver, who had the Airbus wingtip impacted the Boeing while very limited vision to the rear of the aircraft. the captain was transmitting his situation. Both ground crew members had observed the Boeing 757 pass behind their aircraft and turn Although the company had been advised that into the centerline of its parking stand. Both the Boeing was returning to the parking stand

20 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 after the parking guidance lights had been au- left wing. First the left, then the right wing, tomatically activated by a timer, the lack of broke free from the fuselage. There was no available dispatchers at the time caused a de- fire. The occupants all evacuated the wreck- lay in assigning someone to meet the aircraft. age without assistance, and the only reported The collision had occurred by the time this injury was to a passenger who had been in the was accomplished. right cockpit seat.

AAirir T Taxiaxi/ CCommuterommuter The investigation revealed that the trail made by the nosewheel indicated that it had not been centered at the start of the takeoff roll. The pattern of runway dirt thrown up by the wheel indicated misuse of nosewheel steering, which can lead to loss of directional control. There had been no attempt to abort the takeoff either when directional control was initially lost or when the right wing hit the Press on Regardless ground shortly after the first liftoff. Scrape Is Not Prudent Practice marks made by the tail skid indicated that the aircraft was forced to lift off at an airspeed de Havilland DHC-6 Twin Otter: Substantial damage. that was too low to sustain flight and in a One serious injury. direction not aligned with the runway.

The pilot elected to make a short-field takeoff The pilot was cited for misuse of nosewheel from the 2,400-foot gravel runway. Passen- steering early in the takeoff sequence and for gers included four occupants in addition to not aborting the takeoff. the pilot and one large dog. The weather was clear and cold with visibility more than 15 miles. Open Drain Valve Brings Aircraft Down The pilot lost directional control shortly after the takeoff run began. The aircraft veered to McDonnell Douglas DC-6: Aircraft destroyed. Fatal the right and then swung back across the run- injuries to three. way to the left edge along which the pilot attempted to continue the takeoff. The air- The four-engine aircraft was forced to make craft lifted off after a ground run of approxi- an emergency descent because of fuel exhaus- mately 1,200 feet but the right wing dropped tion. The late afternoon cargo flight was on and scraped the ground, after which it touched instrument flight rules (IFR) in visual meteo- down on the main gear with the tail skid drag- rological conditions (VMC). One propeller was ging on the surface. Another change in direc- turning under power while the other three tion aimed the aircraft to the right again, and were windmilling as the aircraft neared the it left the runway at an angle of 35 degrees ground. On landing, the main gear struck the and traveled through brush along the side of top of a levee causing the right wing to hit the the runway. ground below the levee. The aircraft broke apart and the wreckage came to rest in a drainage The pilot continued the attempt to become canal. The three crew members were fatally airborne and the aircraft lifted off again. The injured in the accident. left wing collided with the roof of a wooden shed and continued flying a few feet above The aircraft had departed with seven hours of the ground. After continuing for approximately fuel, 2.7 hours of which had been lost because, 400 feet, the aircraft flew over a 50-foot em- according to the accident report, there was an bankment, then rolled to the left and dropped open drain valve discovered in the wreckage to the ground where it impacted first on the inside the number four engine nacelle. Inves-

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 21 tigators estimated that the valve had been open collided with a parked aircraft which it pushed for an extended time, but no ground person- back into an embankment. nel reported having seen fuel draining from it prior to the flight. There was no fire and the pilot and passenger deplaned through the cabin door without injury. The aircraft sustained damage to the right wing, nose cowling, left propeller and left engine. CCorporateorporate EExecutivexecutive Rushed Pilot Loses Track of Altitude

Beechcraft B55: Substantial damage. Fatal injuries to one.

Engine Stops The aircraft was flying under instrument flight During Taxiing rules (IFR) at night on approach to its destination. There were one crew member and one Beechcraft B55A Baron: Substantial damage. No passenger aboard. The pilot had listened to injuries. the automatic terminal information service (ATIS) and learned that the nondirectional The pilot, with one passenger aboard, had landed beacon (NDB) approach to runway 21 was in after an international flight and had cleared use. customs. The control tower cleared the pilot to taxi from the customs parking area to a As the aircraft approached the NDB, the pilot parking area elsewhere on the airport. was asked by the air traffic controller if he wanted the localizer approach to runway 15 The final portion of the taxi route to the desig- instead of the NDB approach. The pilot chose nated parking area required the pilot to nego- the localizer approach and was given vectors tiate a narrow taxiway that sloped downhill to intercept it. As the pilot intercepted the and had vehicles and aircraft parked closely localizer course, he was asked by air traffic on both sides. Wind speed was calm and the control (ATC) whether he was inside or out- taxiway was dry. Witnesses reported that the side the NDB fix. The pilot looked at the aircraft’s taxi speed was faster than normal distance measuring equipment (DME) but was for the prevailing conditions but the pilot said unable to read it because of the glare from a he was taxiing very slowly. flashlight the passenger was using to prepare charts for landing at an alternate airport if an While the aircraft was travelling on the down- overshoot was required. ward sloping portion of the narrow taxiway, the right engine stopped but because both The pilot then looked at the course deflection throttles were closed, there was no thrust im- indicator (CDI) of the number one very high balance to make the pilot aware that an engine frequency omnidirectional range (VOR) receiver had ceased operation. At the bottom of the to find his position on the localizer. The needle slope, however, the pilot noticed the station- showed a full deflection to the left, and the ary right propeller as the thrust from the oper- pilot replied that he was inside the fix. There ating left engine began to veer the aircraft to was no further contact with the pilot and a the right. Application of left rudder and brake search was initiated minutes later. The aircraft had no effect and the aircraft’s right wing passed was located three and one half hours later, over the fronts of parked automobiles and struck five nautical miles from the threshold of the the roof of a van. The impact caused the air- runway and one-fifth nautical mile west of the craft to swing sharply to the right where it NDB fix.

22 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 The accident report states that the pilot had days, and took off without having filed a flight prepared himself for an NDB approach. While plan; a nearby resident heard the aircraft de- the pilot was being vectored for the localizer part and expressed surprise that someone was approach to runway 15, he was also selecting flying in the low clouds and poor visibility. and analyzing the approach plates, slowing The pilot was rated for instrument flight. the aircraft, descending and lowering the flaps and gear. The report states that the speed of Approximately an hour into the flight the pi- the aircraft rushed the pilot. He entered 79 lot made his first radio contact, which was to degrees into the omni-bearing selector of the an air traffic control (ATC) facility reporting VOR, instead of 75 degrees. When ATC asked that he was IMC (in instrument meteorologi- the pilot if he was past the NDB fix, the pilot cal conditions) without flight information. He stated incorrectly that he was. was given local weather conditions that included rain, 5/8 cloud coverage at 600 feet The pilot would be expected to start his de- and overcast at 800 feet. He was within ap- scent after crossing the fix from 2,000 feet above proximately 30 miles of his intended destina- sea level (asl); the minimum descent altitude tion and switched to the ATC unit serving that is 1,080 feet asl. However, the report states, area for airport and weather information, which because the aircraft was off course to the right, was consistent with the previous weather re- the pilot was probably concentrating on report. There was a one-degree temperature- turning to the localizer and was inattentive to dewpoint spread. altitude. The aircraft flew over a golf course which has an elevation of 920 feet asl and The pilot was given a clearance and performed struck the trees. a VOR approach. He reported that the aircraft had made the turn to base leg as called for by The aircraft was destroyed, the pilot was seri- the procedure. ATC requested that the pilot ously injured and the passenger was fatally report when he had the airport in sight and injured. The pilot was cited for being unsure then enter a downwind leg, and the pilot ac- of his position on the approach; he was un- knowledged. The pilot was advised that the aware that he had descended too low for the cloud base at 600 feet had become overcast approach until the aircraft struck the trees. and reported that he could see the ground but not the airport.

A pilot on the ground reported sighting the OOtherther aircraft between gaps in the clouds northwest GGeneraleneral AAviationviation of the airport at a height he estimated to be between 200 and 300 feet above the ground. ATC advised the pilot of his location and that the aircraft was headed for high ground. The pilot acknowledged and advised he was making a missed approach, next reporting he was passing through 800 feet.

Low Clouds, Poor Visibility, Shortly thereafter, when ATC requested his Disorientation, No Flight Plan intentions, the pilot reported that he had a Produce Expected Result major problem, that he was suffering spatial disorientation and was “in a loop.” This was Cessna 175: Aircraft destroyed. Fatal injuries to the last radio transmission from the aircraft. one. Rescue services were alerted and a helicopter that backtracked the last known course of the The pilot arrived at the closed small airport aircraft found the wreckage of the aircraft in a slightly prior to 0800 hours on the early spring field less than 20 minutes after the final radio day. He had been visiting the area for three transmission — the pilot had been killed.

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 23 A radar plot of the approach revealed that the began to roll down a slope. The main wheels aircraft had initially arrived over the VOR sta- were stopped by a small ditch and the aircraft tion on course but that the approach proce- nosed over, coming to rest inverted. The pas- dure was not flown accurately in either direc- senger was able to evacuate with minor inju- tion or altitude. It had disappeared tempo- ries. The aircraft sustained a broken propeller rarily from radar coverage at one point when and other damage to the wing leading edges, it was northwest of the airport and subsequently fuselage frame, engine cowling air intake and followed an erratic path and made rapid changes windshield. in altitude, consistent with the pilot’s report of spatial disorientation and perception that the aircraft was in a loop.

RotorcraftRotorcraft The aircraft had struck the ground in a near vertical attitude and at a high airspeed; it was demolished on impact and, although little information could be obtained from the wreckage, no evidence was found of pre-impact problems with the structure or mechanisms. The deviation pointer of the VOR system had had a problem but was repaired some time previously. Two Pilots at One Time Results in Hard Landing

Aircraft Taxis on its Own Robinson R22 Beta: Substantial damage. No injuries. Piper J-3 Cub: Extensive damage. Minor injuries to one. The aircraft was en route to a private landing site. Aboard were the pilot and a passenger, The engine of the two-place, tailwheel aircraft both of whom were private pilots with heli- was not equipped with an electric starter and copter ratings; the passenger was the more required the propeller to be swung by hand to experienced pilot in this type aircraft. start it. The passenger for this flight was not a pilot, but had flown with the pilot numerous The pilot decided to make a practice landing times and had been trained by him to assist in in a field prior to arriving at the destination the engine starting procedure to the extent of landing site. After accomplishing the landing holding the control stick back and applying without incident, the pilot elected to execute a the brakes. He had not been briefed, however, running takeoff. The aircraft slewed slightly on operation of the magnetos or on the opera- to the right during the takeoff. The pilot was tion of the throttle. not alarmed at this and was satisfied that the aircraft was under control; however, the pas- The pilot pumped the throttle and set it to the senger considered that the situation warranted starting position, switched on the magnetos that he take control. For a short period, both and went in front of the aircraft to swing the pilots attempted to fly the aircraft. The pilot propeller by hand. The passenger held the stated that he did not hand over control of the control column back and applied brake pres- aircraft to the passenger. sure. The engine started on the first swing of the propeller, but operated at a much higher During the confusion, the aircraft made a hard rpm than normal. landing. The tail boom and cabin structure were wrinkled and the landing skids splayed The passenger was unable to hold the aircraft outward. There was no fire and no injuries stationary with the brakes and did not know were sustained by the occupants who exited how to reduce the engine power. The aircraft without further incident.

24 F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 Rocks Hide in occluded front. The pilot elected to continue Low-lying Mists the flight to the destination and, under conditions of very little visibility, he descended to a Messerchmitt-Bölkow Blohm BK117: Aircraft de- very low altitude. The helicopter collided with stroyed. Fatal injuries to 10. a hill covered by fog and mist. The aircraft was destroyed by the impact and all 10 occu- The rotorcraft had departed during the early pants sustained fatal injuries. evening in visual meteorological conditions (VMC) with a crew of two and eight passen- Factors cited in the accident included incor- gers aboard. rect pilot decisions by not maintaining VMC conditions after the aircraft encountered re- During the flight, the weather deteriorated strictions to vision, and poor monitoring by because of an approaching typhoon and an the flight service operator. o

F LI GHT SAFETY FOUNDATI O N • F LI GH T SA F E TY D I G E S T • MARCH 1992 25